Key facts
- Japan will "actively and continuously review" AI-related laws.
- The goal is to enhance the country's ability to respond to threats from high-performance AI models.
- Anthropic's Claude Mythos is mentioned as an example of a powerful AI model.
- The government's new AI policy emphasizes a risk-based, agile, and multistakeholder approach.
- Japan aims to balance AI risks with beneficial innovation and adoption.
Japan is set to "actively and continuously review" its AI-related laws to bolster its defenses against threats posed by advanced artificial intelligence models, such as Anthropic's Claude Mythos. This strategic move, outlined in a new government AI policy released on Thursday, signals a commitment to a dynamic regulatory framework.
The policy emphasizes a risk-based, agile, and multistakeholder approach to AI governance, aiming to strike a balance between mitigating potential dangers and fostering beneficial innovation. This contrasts with more rigid, one-size-fits-all regulatory models seen elsewhere, such as the European Union's draft Artificial Intelligence Act.
Globally, governments are grappling with AI regulation, with countries like the U.S., Canada, and the EU developing specific legislation. Japan's approach, rooted in its 2019 Social Principles of Human-Centric AI, prioritizes maximizing AI's positive societal impact through principles like human dignity, diversity, sustainability, privacy protection, security, fair competition, fairness, accountability, transparency, and innovation.